Over the years, LDRD at Berkeley Lab has provided early investments for exciting new research efforts that have grown into major research programs, including:
- Open Molecules 2025, an unprecedented dataset of molecular simulations to train machine learning tools that can more quickly and accurately model molecular motion and chemical reactions;
- A new technique to improve quantum sensing, involving encasing nanodiamonds in tiny moving droplets of water to detect trace amounts of certain ions and molecules; and
- Development of a new kind of automated lab, called A-Lab, that uses robots guided by artificial intelligence to accelerate the development of useful new materials.
“Many of the scientific challenges our nation is facing require this kind of out-of-the-box thinking that can lead to high impact results,” said Berkeley Lab Chief Research Officer Carol Burns.
Today, early LDRD investments will jumpstart valuable new research directions in AI, quantum systems, microelectronics, fusion energy, and the bioeconomy – which will further America’s competitive edge as a world scientific and technological leader.
In addition to seeding promising new research, the program also supports the innovative, risk-taking scientists and engineers behind the work. Across three funding tracks – Multi-Area Topics, Early Career Development, and Area Priorities – the program invests in scientists probing creative new concepts and methods that otherwise may struggle to find funding.
“Unlike most government funding that favors projects that have already had some early results or success, LDRD funding allows you to explore out-there ideas,” said Kristin Persson, Director of the Materials Project at Berkeley Lab. The Materials Project started as an internal LDRD project and now serves more than 650,000 registered users worldwide and is a core Department of Energy program for designing novel materials for everything from batteries to computer chips.
The FY26 cohort of LDRD projects tackles the frontiers of scientific discovery head on. From 170 submissions, 113 projects have been funded at a current total value of $25.7 million.
Here are five select, newly seeded projects that showcase the breadth of exploration at Berkeley Lab:
Understanding & Predicting Power & Energy Usage in Large-Scale HPC/AI Models – Won Young Park
Extensive research has been conducted on data center power consumption along with newer technology trends. However, it remains challenging to fully understand and predict the energy behavior of AI applications and implement local power management strategies to save energy while meeting application power requirements. This project will build a base prediction model for the energy and performance of AI applications, incorporating performance trade-offs associated with various power management strategies.
Biological recovery of rare-earth elements (REEs) – Ning Sun
This project aims to discover microbial strains that recover REEs from electronic wastes, and ultimately develop a cost-effective bioprocess to achieve a REE circular economy.
Machine Learning Enabled Novel Discovery for PFAS Degradation – Romy Chakraborty
To accelerate the search for solutions to “forever chemical” PFAS contamination, a new pipeline is being pioneered that applies artificial intelligence and machine learning to microbial datasets in order to identify key functional features that enable PFAS degradation. From there, the team will downselect the most promising microbes capable of efficiently breaking down these persistent chemicals.
Development of superconducting film for sensing and quantum applications – Aritoki Suzuki
This project brings together the materials science and nanofabrication expertise of the Molecular Foundry with the sensor development expertise of the lab’s Physics Division to explore how microfabrication processes influence materials properties, ultimately, device performance.
Manipulating materials and their properties within confined geometry using advanced nanofabrication methodologies – Qi Zhang
The sub-10 nm regime is a critical threshold where materials and devices exhibit unique properties that differ significantly from their bulk counterparts. The advancement of techniques such as extreme ultraviolet (EUV) lithography offers unique opportunities not only to scale microelectronic devices down to sub-10 nm dimensions but also to manipulate and understand materials at this scale. Zhang aims to develop the methodology to construct a two-dimensional (2D) metal/semiconductor lateral structure at sub-10 nm scale using advanced nanofabrication techniques.
The future of LDRD exploration continues with the FY27 call for proposals issued on November 19. The deadline for multi-area proposals will be March 13, 2026, while Area Priority and Early Career Development proposals are due on April 17. Visit the LDRD website for more information about the LDRD call for proposals.
